lljyangfan at gmail.com
Thu Jun 20 06:01:46 EDT 2013
Thank you for your answer.
That is to say the designer use __USER_DS here deliberately to improve the
efficiency jump from kernel space to user space?
BTW,how can you find this email? I write this email On 10/24/2012.
2013/6/20 Baoquan He <baoquan.he at gmail.com>
> On 10/24/2012 08:04 PM, Fan Yang wrote:
> > Hi all: I print the cs ds and ss register in the user space, and it
> > is same as the __USER_CS and __USER_DS which defined in kernel as 73
> > and 7b. In the kernel __KERNEL_CS and __KERNEL_DS defined as 60 and
> > 68, but when I print this two value in my kernel module, I get 60
> > and 7b. Why ? It should be 60 and 68, shouldn't it?
> Hi Fan,
> I just talked to you and very impressed. By checking source code, I
> found what you mentioned is normal.
> You can check arch/x86/kernel/entry_32.S, and go to the page_fault
> definition, in error_code, you can
> see the CS and DS assignment.
> /* the function address is in %gs's slot on the stack */
> pushl_cfi %fs
> /*CFI_REL_OFFSET fs, 0*/
> pushl_cfi %es
> /*CFI_REL_OFFSET es, 0*/
> pushl_cfi %ds
> /*CFI_REL_OFFSET ds, 0*/
> pushl_cfi %eax
> CFI_REL_OFFSET eax, 0
> pushl_cfi %ebp
> CFI_REL_OFFSET ebp, 0
> pushl_cfi %edi
> CFI_REL_OFFSET edi, 0
> pushl_cfi %esi
> CFI_REL_OFFSET esi, 0
> pushl_cfi %edx
> CFI_REL_OFFSET edx, 0
> pushl_cfi %ecx
> CFI_REL_OFFSET ecx, 0
> pushl_cfi %ebx
> CFI_REL_OFFSET ebx, 0
> movl $(__KERNEL_PERCPU), %ecx
> movl %ecx, %fs
> GS_TO_REG %ecx
> movl PT_GS(%esp), %edi # get the function address
> movl PT_ORIG_EAX(%esp), %edx # get the error code
> movl $-1, PT_ORIG_EAX(%esp) # no syscall to restart
> REG_TO_PTGS %ecx
> SET_KERNEL_GS %ecx
> Below is the assignment. This is changed in 2.6, the reason is that in
> kernel the CPL is 0 and it's
> safe to operate __USER_DS with DPL is 3. Here if use __KERNEL_DS, even
> though they have the
> same content, it need change back when return to user space.
> So in your original post, you mentioned the cs:ds is 60, 7b, it's normal
> and correct value. Hope
> this can help.
> movl $(__USER_DS), %ecx
> movl %ecx, %ds
> movl %ecx, %es
> movl %esp,%eax # pt_regs pointer
> call *%edi
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Kernelnewbies